The effectiveness of programs designed to protect against cancer is likely to be greatly improved by the identification of risk factors that are related to exposures and reactions at the cellular level. Identifying serologic precursors is one step in this direction. Nutritionally related substances such as certain vitamins, trace metals, and lipid-related factors appear well worth further investigation. The major source of serum for this study came from a program in 1974 in Washington County, Maryland, in which 25,802 persons each donated 15 ml of blood. The serum has been kept frozen at -73 degrees C since that time. Cancer cases developing among the serum donors have been identified by the Washington County cancer register maintained by our center. By January 1, 1986, 1386 cases among serum donors had been registered. By July 1, 1987, pre- diagnostic serum levels of retinol, retinol-binding protein, beta- carotene, alpha-tocopherol, selenium, and cholesterol will have been assayed for cancers of the lung, colon, rectum, skin, bladder, pancreas, and prostate, along with sera from two controls matched to each case. It is proposed to extend these assays to melanomas and to cancers of the breast, cervix, endometrium, kidney, mouth and pharynx, and ovary, again with two controls matched to each case. In addition to the serum components assayed previously, we propose to add total carotenoids, zinc, and apoproteins. Interactions between serum components will be examined with special reference to case-control differences. Comparisons of pre- and post-diagnostic levels of the preceding serum components will be made among cases for whom both of these specimens are available. Case-control comparisons will be made for cases whose cancers were diagnosed prior to giving blood in 1974. Findings from these two comparisons will be helpful in interpreting the results of the usual case-control studies in which sera are assayed among recently diagnosed cases and their controls.